Summary of Key Points
The AI industry has moved from a capital-driven phase focused on "spending money on technology" to an industry-driven phase aimed at generating revenue, entering the second half of its development where applications are being scaled up and commercialization is accelerating. The focus of investment and financing has shifted from upstream technologies to downstream use cases, and the path for companies to go public has become clearer (with the STAR Market providing a favorable environment for large-model companies). Short-term adjustments in the stock market represent excellent opportunities for strategic positioning, and the autumn market is expected to be promising. In the future, investments should focus on core AI assets and related areas with potential for growth.
I. The AI Industry Enters the “Second Half”: From “Building the Infrastructure” to “Putting It into Use”
Previously, AI was like building the infrastructure—everyone was working on foundational technologies such as large models and computing power (e.g., ChatGPT, GPU chips). Now, we have entered the phase of actually applying AI in practical scenarios, such as AI software, robotics, and autonomous driving, with the goal of generating revenue. Zhang Yidong, Chief Economist at Haitong International, made an analogy: AI is similar to the internet in 1998; back then, the penetration rate in the US was just 30%, but by 2026, it had risen to 28.3% and is set to accelerate further. In the future, AI will permeate all industries, with applications like using AI robots in factories and AI-assisted diagnostics for doctors, leading to a variety of new business models.
II. A Shift in Investment and Financing: Downstream Applications Become the Priority
In the past, investments in AI were based on the quality of the technology; now, the focus is on whether it can generate profit. Lu Ying, Director of the Research Institute at Guotai Haitong, noted that funds are primarily flowing towards downstream application services (e.g., AI software, robotics), indicating a shift from focusing on potential applications to generating actual revenue. Data supports this trend: in the first quarter of 2026, there were 1,332 AI-related financings in China, a year-on-year increase of 745, with AI software applications (24.2%), embodied intelligence (12.9%), and autonomous driving (8.6%) leading the list. As underlying large models mature, the cost of developing new products is lower, and the pace of innovation is faster, making financing for downstream applications particularly popular.
III. Accelerated Listing of AI Companies: Clear Path and Right Time
Previously, there were no clear criteria for AI large-model companies to go public. However, the Shanghai Stock Exchange has released the fifth set of guidelines for the STAR Market, which also includes robotics within the scope of high-end equipment listings, effectively creating a “green channel” for these companies. Lu Ying believes this initiative will facilitate a cycle of investment and exit in the primary market, benefiting the entire industry. In the first quarter of 2026, 23 AI companies went public, 18 more than in the same period last year, largely due to strong sales of hardware (e.g., robotics, AI equipment) and faster review processes. Yu Weijun from Guotai Haitong advises that companies should align with national strategic priorities, prepare for an IPO during their growth phase, achieve rapid revenue growth while reducing losses, and maintain a cash reserve of 18-24 months to avoid financial shortages.
IV. Stock Market Adjustments as Opportunities: The Autumn Market Is Approaching
There are concerns about an oversupply of AI computing power, leading to a pullback in tech stocks. However, Zhang Yidong sees this as an opportunity within a crisis. He believes the Chinese stock market will experience a long-term upward trend for several reasons:
1. AI is driving growth in high-tech industries (the added value of high-tech sectors increased by 14.1% year-on-year in June 2026, 8 percentage points higher than the overall manufacturing sector).
2. AI is boosting exports (for example, a significant increase in integrated circuit exports).
3. Funds are expected to flow back into the market (as residents shift their investments to stocks and overseas capital returns).
He predicts that after adjustments in June or July, the market will rebound in the autumn, with AI-related trends spreading across more sectors.
V. Future Investment Directions: Focus on “Core Assets”
Zhang Yidong recommends using a SMART strategy for asset selection, focusing on three key areas:
1. High-tech Hardware: Core assets in short supply for AI (e.g., computing power chips, large models).
2. Security Assets: Resources such as copper and rare earths, as well as energy (since AI requires substantial amounts of electricity).
3. Manufacturing for Overseas Markets: Industries that can improve efficiency with AI and sell products internationally (e.g., power equipment, chemicals).
Additionally, securities firms with a high presence in AI-related services (e.g., those assisting AI companies in going public) are also worth considering.
In summary, AI is no longer just a concept; it is a real industry that can generate profit and drive economic growth. Both businesses and investors should seize the opportunities presented in this “second half” of the AI development phase.